Biology subjects

Silveira, M.

Publications and source records attributed to Silveira, M..

3 recordsLinked to original sources

bioRxiv2023

3DVascNet: an automated software for segmentation and quantification of vascular networks in 3D

BackgroundAnalysis of vascular networks is an essential step to unravel the mechanisms regulating the physiological and pathological organization of blood vessels. So far, most of the analyses are performed using 2D projections of 3D networks, a strategy that has several obvious shortcomings. For instance, it does not capture the true geometry of the vasculature, and generates artifacts on vessel connectivity. These limitations are accepted in the field because manual analysis of 3D vascular networks is a laborious and complex process that is often prohibitive for large volumes. MethodsTo overcome these issues, we developed 3DVascNet, a deep learning (DL) based software for automated segmentation and quantification of 3D retinal vascular networks. 3DVascNet performs segmentation based on a DL model, and it quantifies vascular morphometric parameters such as the vessel density, branch length, vessel radius, and branching point density. ResultsWe tested 3DVascNets performance using a large dataset of 3D microscopy images of mouse retinal blood vessels. We demonstrated that 3DVascNet efficiently segments vascular networks in 3D, and that vascular morphometric parameters capture phenotypes detected by using manual segmentation and quantification in 2D. In addition, we showed that, despite being trained on retinal images, 3DVascNet has high generalization capability and successfully segments images originating from other datasets and organs. More-over, the source code of 3DVascNet is publicly available, thus it can be easily extended for the analysis of other 3D vascular networks by other users. ConclusionsOverall, we present 3DVascNet, a freely-available software that includes a user-friendly graphical interface for researchers with no program-ming experience, which will greatly facilitate the ability to study vascular networks in 3D in health and disease.

bioinformatics↗

bioRxiv2023

Biodiversity responses to forest cover loss: taxonomy and metrics matter

The actions required for biodiversity conservation depend on species responses to habitat loss, which may be either neutral, linear, or non-linear. Here, we tested how taxonomic, functional, and phylogenetic diversity of aquatic insects, dragonflies, frogs, and terrestrial mammals, as well as their species composition respond to forest cover loss. We hypothesized that taxonomic, functional, and phylogenetic diversity would respond nonlinearly (thresholds) to forest cover loss. Our findings do not support the current idea that a single threshold value of forest cover loss is applicable across tropical regions, or that some biodiversity facets are consistently more sensitive than others across different taxa. Species compositional responses to forest cover loss showed general patterns with thresholds between 30-50%. These results highlight the importance to consider multiple biodiversity facets when assessing the effects of forest cover loss on biological communities.

ecology↗

bioRxiv2022

Detection and measurement of butterfly eyespot and spot patterns using convolutional neural networks

Butterflies are increasingly becoming model insects where basic questions surrounding the diversity of their color patterns are being investigated. Some of these color patterns consist of simple spots and eyespots. To accelerate the pace of research surrounding these discrete and circular pattern elements we developed two distinct convolutional neural networks (CNNs) for detection and measurement of butterfly spots and eyespots on digital images of butterfly wings. We tested the accuracy of the detection and of the area measurements using manual identifications and measurements. These methods were able to identify and distinguish marginal eyespots from spots, as well as distinguish these patterns from less symmetrical patches of color. In addition, the measurements of an eyespots central area and surrounding rings were highly accurate. These CNNs offer improvements of eyespot/spot detection and measurements relative to previous methods because it is not necessary to mathematically define the feature of interest. All that is needed is to point out the images that have those features to train the CNN. Author summaryWe developed two distinct convolutional neural networks (CNNs) for detection and measurement of butterfly spots and eyespots on digital images of butterfly wings. We tested the accuracy of the detection and of the area measurements using manual identifications and measurements. Our methods were able to identify and distinguish marginal eyespots from spots, as well as distinguish these patterns from less symmetrical patches of color. In addition, the measurements of an eyespots central area and surrounding rings were highly accurate.

systems biology↗